Add ExcServerSuccessfulReturnValue() and its test

Add ExcServerSuccessfulReturnValue() and its test.

There’s also some light reformatting in here. Should save a few lines.

TEST=util_test ExcServerVariants.ExcServerSuccessfulReturnValue
R=rsesek@chromium.org

Committed: https://chromium.googlesource.com/crashpad/crashpad/+/57eb311528f23c40d0120e4c6c2ef3a1220c863d

[Mark Mentovai, 2014-09-16 20:28:31]

mark@chromium.org changed reviewers:
+ rsesek@chromium.org

[Mark Mentovai, 2014-09-16 20:28:31]

Here’s something else that I was just going to wind up duplicating in another
file, so I refactored it into a common utility function.

This depends on https://codereview.chromium.org/565793005/.

[Robert Sesek, 2014-09-17 15:29:39]

https://codereview.chromium.org/574753003/diff/1/util/mach/exc_server_variant...
File util/mach/exc_server_variants.cc (left):

https://codereview.chromium.org/574753003/diff/1/util/mach/exc_server_variant...
util/mach/exc_server_variants.cc:204: : MachMessageServer::Interface(),
Not sure these clang-format changes are improvements.

https://codereview.chromium.org/574753003/diff/1/util/mach/exc_server_variants.h
File util/mach/exc_server_variants.h (right):

https://codereview.chromium.org/574753003/diff/1/util/mach/exc_server_variant...
util/mach/exc_server_variants.h:442: //! exception handler at a wider (task or
host) scope. This may also be
Do you ever want to let the kernel search for a wider scope?

[Mark Mentovai, 2014-09-17 15:35:38]

https://codereview.chromium.org/574753003/diff/1/util/mach/exc_server_variant...
File util/mach/exc_server_variants.cc (left):

https://codereview.chromium.org/574753003/diff/1/util/mach/exc_server_variant...
util/mach/exc_server_variants.cc:204: : MachMessageServer::Interface(),
rsesek wrote:
> Not sure these clang-format changes are improvements.

I’d say it’s 50/50. What it did to the “check” calls is fine, but the
initializers are less good.

https://codereview.chromium.org/574753003/diff/1/util/mach/exc_server_variants.h
File util/mach/exc_server_variants.h (right):

https://codereview.chromium.org/574753003/diff/1/util/mach/exc_server_variant...
util/mach/exc_server_variants.h:442: //! exception handler at a wider (task or
host) scope. This may also be
rsesek wrote:
> Do you ever want to let the kernel search for a wider scope?

In general, yes, if you decide that you don’t want to handle an exception for
some reason.

If you’re a hardware fault exception catcher (like EXC_BAD_ACCESS) and you don’t
want to handle a bad access that you receive, you want to let the kernel search
for the next exception handler—typically the host-level handler in the kernel
that will turn it into a signal.

[Robert Sesek, 2014-09-17 16:05:33]

LGTM

[Mark Mentovai, 2014-09-17 16:08:23]

Committed patchset #2 (id:20001) manually as 57eb311 (presubmit successful).